The MOONGLOW did not keep a listening watch on VHF channel 5A. Although it is not mandatory for a vessel of the size and type of the MOONGLOW to participate in the VTS reporting system, had she been participating, she would have been aware of the movements of other vessels in the area, including those of the submarine. Additionally, as a result of the non-participation of the MOONGLOW, the submarine and other vessels were not aware of her presence nor of her movements. The operator of the MOONGLOW was steering by magnetic compass. Although the compass had been purchased new less than one year before, the vessel had not been swung to determine the deviation of the compass. The value of deviation on compass headings was unknown as was the accuracy of the courses reported to have been steered. The practice on board was to compare the course shown by the uncompensated magnetic steering compass to the true course to steer displayed by the LoranC. The course to steer displayed by the LoranC is electronically calculated from the vessel's current position and the next previously determined waypoint; in this case, a position south of Race Rocks. The operator of the MOONGLOW reported that, at about 2130, he altered course to 060(C), about 083(T), for a position north of the traffic lane and south of Race Rocks. If this course to steer is laid off on the chart from the intended waypoint back toward a position off Sheringham Point, it would indicate that the vessel was considerably further to the south than the operator stated; south of the outbound traffic lane in US Navy Exercise Area 2. Conversely, if the vessel followed this course when laid off from the 2130 estimated position, the course would not lead to the intended waypoint but on to the shore north of Beechey Head. The operator of the MOONGLOW reported that the collision took place north of the traffic lane because: the approximate track of the vessel took her to about four cables north of the limits of the outbound traffic lane at her course alteration south of Sheringham Point; and the radar was on the four-mile range and Sheringham Point was showing just inside the edge of the screen. The operator of the vessel had not, however, plotted a position on the chart for some 20 minutes prior to the time of the collision. The collision position given by the MOONGLOW cannot be reconciled either with that given by the THOMSON or with the evidence of Seattle Traffic. The collision position recorded by the submarine is about five miles SSW of Sheringham Point. In the half-hour preceeding the collision, the submarine had obtained eight positions by GPS or by radar. These positions place the submarine at or near the centre of the traffic lane. VTS Seattle radar placed the submarine in the traffic lane, i.e. more than 4.4miles from Sheringham Point. While there is no VTS record of the submarine's positions in this period, the vessel presented a good radar target to VTS and the vessel was closely monitored. Had the vessel left the traffic lane, it would have been evident on VTS radar. It is the Sector Operator's duty to warn a vessel if she leaves a traffic lane. No such warning was necessary or given. The traffic lane limits are superimposed electronically on the VTS radar screens and these lines are accurate to within 200 to 300yards (182 to 274m). The Sector Operator would not call another vessel if she were close to the line as in the case of the THOMSON. There can be little doubt that the submarine was at or near the centre of the outbound traffic lane when the collision occurred. The vessel plotted her position on the chart almost continuously. The relative positions of the ketch and the submarine before the collision also tend to confirm that the MOONGLOW was further to the south than her operator believed. Because the submarine, while steering 282(T), sighted the ketch's green light to port, it follows that the ketch, steering a course of about 083(T), must have been in the traffic lane. The relative position is substantiated by the fact that the MOONGLOW sighted the submarine 15 on her starboard bow. Reportedly, the operator of the MOONGLOW sent the MAYDAY immediately after the collision. Because both Canadian and US Coast Guard radio stations recorded this time as 2119, it is most likely to be correct. Although this time is not the same as the times reported by both vessels, it does not appreciably change the collision position as recorded by the THOMSON because the submarine's GPS position of 2126 and a position recorded at 2117 by radar were very close. Both were near the centre line of the traffic lane. The collision occurred 3hours and 30minutes before the predicted time of high water at Sooke, and 15minutes before the predicted time at which the current turned to ebb. The rate of the tidal current was zero. Since departing Nanoose, the submarine was tracked on radar first by Vancouver VTS and then by Seattle VTS, both of which reported that the THOMSON presented a good radar target. The submarine had no anti-radar detection devices on board. The raising of the masts reportedly increased her reflective surface by 50percent. The operator of the MOONGLOW did not see the radar echo of the submarine until she was approximately 1/16 of a mile off, in spite of the fact that the radar was readily in view from the cockpit steering position. However, because the operator was reefing the sails and out of the cockpit for at least part of the submarine's approach, the echo may have been detected by radar but not seen by the operator. While reefing the sail and carrying out other duties requiring his absence from the cockpit, the operator engaged the autopilot. The reason why the MOONGLOW was not detected by the THOMSON nor by Seattle VTS is unknown; the ketch was equipped with a radar reflector and, although constructed of wood, had other reflective surfaces on her rigging. Navigating in the Vicinity of Submarines The yearly edition of the Canadian Notices to Mariners warns that, in restricted waters, submarines should be passed with caution observing their limited manoeuvrability on the surface, deep draught and vulnerability to collision. The bridge watch officer on the THOMSON came hard-to-port in an attempt to avoid collision. With a collision imminent, it is imperative for a submarine of the type of the THOMSON, with a single-skinned hull, to take the impact on the bow to avoid pressure hull damage amidships which could prove fatal to the vessel and her crew. The operator of the MOONGLOW did not hear the fog signals sounded by the submarine and the submarine did not hear a signal from the ketch. The operator of the MOONGLOW stated that he would have sounded the fog horn at least within five minutes prior to the collision. This testimony differs from his declaration to the commander of the THOMSON. At that time, the operator of the MOONGLOW stated that he only emitted sound signals when passing vessels and that he was not doing so at the time of the collision. The COLREGS require that a sailing vessel sound three blasts in succession, namely one prolonged followed by two short blasts, at intervals of not more than two minutes. Because the MOONGLOW was not detected by other means, the sound signal was the only means by which she could have been identified as a vessel in some way restricted in her ability to manoeuvre. The difference in testimony regarding whether the MOONGLOW was under power or under sail cannot be resolved. As there can be little doubt concerning the submarine's position, the collision occurred in the outbound traffic lane. Because the COLREGS require that a vessel of less than 20 metres in length or a sailing vessel shall not impede the safe passage of a power-driven vessel following a traffic lane, the question of whether the MOONGLOW was, or was not, under sail is not an issue.Analysis The MOONGLOW did not keep a listening watch on VHF channel 5A. Although it is not mandatory for a vessel of the size and type of the MOONGLOW to participate in the VTS reporting system, had she been participating, she would have been aware of the movements of other vessels in the area, including those of the submarine. Additionally, as a result of the non-participation of the MOONGLOW, the submarine and other vessels were not aware of her presence nor of her movements. The operator of the MOONGLOW was steering by magnetic compass. Although the compass had been purchased new less than one year before, the vessel had not been swung to determine the deviation of the compass. The value of deviation on compass headings was unknown as was the accuracy of the courses reported to have been steered. The practice on board was to compare the course shown by the uncompensated magnetic steering compass to the true course to steer displayed by the LoranC. The course to steer displayed by the LoranC is electronically calculated from the vessel's current position and the next previously determined waypoint; in this case, a position south of Race Rocks. The operator of the MOONGLOW reported that, at about 2130, he altered course to 060(C), about 083(T), for a position north of the traffic lane and south of Race Rocks. If this course to steer is laid off on the chart from the intended waypoint back toward a position off Sheringham Point, it would indicate that the vessel was considerably further to the south than the operator stated; south of the outbound traffic lane in US Navy Exercise Area 2. Conversely, if the vessel followed this course when laid off from the 2130 estimated position, the course would not lead to the intended waypoint but on to the shore north of Beechey Head. The operator of the MOONGLOW reported that the collision took place north of the traffic lane because: the approximate track of the vessel took her to about four cables north of the limits of the outbound traffic lane at her course alteration south of Sheringham Point; and the radar was on the four-mile range and Sheringham Point was showing just inside the edge of the screen. The operator of the vessel had not, however, plotted a position on the chart for some 20 minutes prior to the time of the collision. The collision position given by the MOONGLOW cannot be reconciled either with that given by the THOMSON or with the evidence of Seattle Traffic. The collision position recorded by the submarine is about five miles SSW of Sheringham Point. In the half-hour preceeding the collision, the submarine had obtained eight positions by GPS or by radar. These positions place the submarine at or near the centre of the traffic lane. VTS Seattle radar placed the submarine in the traffic lane, i.e. more than 4.4miles from Sheringham Point. While there is no VTS record of the submarine's positions in this period, the vessel presented a good radar target to VTS and the vessel was closely monitored. Had the vessel left the traffic lane, it would have been evident on VTS radar. It is the Sector Operator's duty to warn a vessel if she leaves a traffic lane. No such warning was necessary or given. The traffic lane limits are superimposed electronically on the VTS radar screens and these lines are accurate to within 200 to 300yards (182 to 274m). The Sector Operator would not call another vessel if she were close to the line as in the case of the THOMSON. There can be little doubt that the submarine was at or near the centre of the outbound traffic lane when the collision occurred. The vessel plotted her position on the chart almost continuously. The relative positions of the ketch and the submarine before the collision also tend to confirm that the MOONGLOW was further to the south than her operator believed. Because the submarine, while steering 282(T), sighted the ketch's green light to port, it follows that the ketch, steering a course of about 083(T), must have been in the traffic lane. The relative position is substantiated by the fact that the MOONGLOW sighted the submarine 15 on her starboard bow. Reportedly, the operator of the MOONGLOW sent the MAYDAY immediately after the collision. Because both Canadian and US Coast Guard radio stations recorded this time as 2119, it is most likely to be correct. Although this time is not the same as the times reported by both vessels, it does not appreciably change the collision position as recorded by the THOMSON because the submarine's GPS position of 2126 and a position recorded at 2117 by radar were very close. Both were near the centre line of the traffic lane. The collision occurred 3hours and 30minutes before the predicted time of high water at Sooke, and 15minutes before the predicted time at which the current turned to ebb. The rate of the tidal current was zero. Since departing Nanoose, the submarine was tracked on radar first by Vancouver VTS and then by Seattle VTS, both of which reported that the THOMSON presented a good radar target. The submarine had no anti-radar detection devices on board. The raising of the masts reportedly increased her reflective surface by 50percent. The operator of the MOONGLOW did not see the radar echo of the submarine until she was approximately 1/16 of a mile off, in spite of the fact that the radar was readily in view from the cockpit steering position. However, because the operator was reefing the sails and out of the cockpit for at least part of the submarine's approach, the echo may have been detected by radar but not seen by the operator. While reefing the sail and carrying out other duties requiring his absence from the cockpit, the operator engaged the autopilot. The reason why the MOONGLOW was not detected by the THOMSON nor by Seattle VTS is unknown; the ketch was equipped with a radar reflector and, although constructed of wood, had other reflective surfaces on her rigging. Navigating in the Vicinity of Submarines The yearly edition of the Canadian Notices to Mariners warns that, in restricted waters, submarines should be passed with caution observing their limited manoeuvrability on the surface, deep draught and vulnerability to collision. The bridge watch officer on the THOMSON came hard-to-port in an attempt to avoid collision. With a collision imminent, it is imperative for a submarine of the type of the THOMSON, with a single-skinned hull, to take the impact on the bow to avoid pressure hull damage amidships which could prove fatal to the vessel and her crew. The operator of the MOONGLOW did not hear the fog signals sounded by the submarine and the submarine did not hear a signal from the ketch. The operator of the MOONGLOW stated that he would have sounded the fog horn at least within five minutes prior to the collision. This testimony differs from his declaration to the commander of the THOMSON. At that time, the operator of the MOONGLOW stated that he only emitted sound signals when passing vessels and that he was not doing so at the time of the collision. The COLREGS require that a sailing vessel sound three blasts in succession, namely one prolonged followed by two short blasts, at intervals of not more than two minutes. Because the MOONGLOW was not detected by other means, the sound signal was the only means by which she could have been identified as a vessel in some way restricted in her ability to manoeuvre. The difference in testimony regarding whether the MOONGLOW was under power or under sail cannot be resolved. As there can be little doubt concerning the submarine's position, the collision occurred in the outbound traffic lane. Because the COLREGS require that a vessel of less than 20 metres in length or a sailing vessel shall not impede the safe passage of a power-driven vessel following a traffic lane, the question of whether the MOONGLOW was, or was not, under sail is not an issue. The collision occurred in dense fog near the centre line of the outward bound traffic lane. Neither vessel was aware of the other's presence until moments before the collision. The MOONGLOW was being operated single-handed. It is unknown whether the MOONGLOW was proceeding under sail or power. The MOONGLOW was not participating in the Seattle Traffic (VTS) Reporting System nor was she required to by regulation. The MOONGLOW was not detected by VTS radar. The steering compass on the MOONGLOW had not been swung, compass deviation was unknown, and the true courses steered by the ketch are not known with certainty. Reportedly, the operator of the MOONGLOW did not plot positions on the chart at regular intervals to monitor the vessel's progress. The MOONGLOW inadvertently strayed into the outbound traffic lane and was transiting the traffic lane in the wrong direction at a small angle to the direction of flow. The MOONGLOW did not observe the radar echo of the THOMSON until she was 1/16 of a mile distant. It is likely that the operator of the MOONGLOW did not observe the presence of the submarine on radar earlier because he had been preoccupied with reefing the main sail prior to the collision. The MOONGLOW had reportedly sounded the prescribed fog signal within five minutes prior to the collision. The MOONGLOW did not hear the sound signals reportedly made by the THOMSON. Upon sighting the THOMSON, the operator of the MOONGLOW instinctively moved to prevent personal injury and did not attempt an emergency manoeuvre to avoid collision. The THOMSON participated fully in the Seattle Traffic Reporting System prior to the collision. The THOMSON presented a good radar target to VTS radar during her outward bound passage. VTS radar tracked the THOMSON continuously. The VTS Sector Operator reported that, at the time of the collision, the submarine was within the outbound traffic lane. Between 2039 and 2126, the THOMSON plotted her position on the chart ten times; at no time during that period was the submarine outside the outbound traffic lane. The THOMSON did not hear the sound signal reported to have been made by the MOONGLOW within five minutes of the collision. It is unknown why the THOMSON did not observe the MOONGLOW on any of her three operational and manned radars. The THOMSON did not acknowledge that a collision had occurred until almost two hours after the collision.Findings The collision occurred in dense fog near the centre line of the outward bound traffic lane. Neither vessel was aware of the other's presence until moments before the collision. The MOONGLOW was being operated single-handed. It is unknown whether the MOONGLOW was proceeding under sail or power. The MOONGLOW was not participating in the Seattle Traffic (VTS) Reporting System nor was she required to by regulation. The MOONGLOW was not detected by VTS radar. The steering compass on the MOONGLOW had not been swung, compass deviation was unknown, and the true courses steered by the ketch are not known with certainty. Reportedly, the operator of the MOONGLOW did not plot positions on the chart at regular intervals to monitor the vessel's progress. The MOONGLOW inadvertently strayed into the outbound traffic lane and was transiting the traffic lane in the wrong direction at a small angle to the direction of flow. The MOONGLOW did not observe the radar echo of the THOMSON until she was 1/16 of a mile distant. It is likely that the operator of the MOONGLOW did not observe the presence of the submarine on radar earlier because he had been preoccupied with reefing the main sail prior to the collision. The MOONGLOW had reportedly sounded the prescribed fog signal within five minutes prior to the collision. The MOONGLOW did not hear the sound signals reportedly made by the THOMSON. Upon sighting the THOMSON, the operator of the MOONGLOW instinctively moved to prevent personal injury and did not attempt an emergency manoeuvre to avoid collision. The THOMSON participated fully in the Seattle Traffic Reporting System prior to the collision. The THOMSON presented a good radar target to VTS radar during her outward bound passage. VTS radar tracked the THOMSON continuously. The VTS Sector Operator reported that, at the time of the collision, the submarine was within the outbound traffic lane. Between 2039 and 2126, the THOMSON plotted her position on the chart ten times; at no time during that period was the submarine outside the outbound traffic lane. The THOMSON did not hear the sound signal reported to have been made by the MOONGLOW within five minutes of the collision. It is unknown why the THOMSON did not observe the MOONGLOW on any of her three operational and manned radars. The THOMSON did not acknowledge that a collision had occurred until almost two hours after the collision. The collision occurred because the operator of the MOONGLOW did not plot his vessel's position frequently to monitor the vessel's progress along her intended track. As a result, the ketch inadvertently strayed into the outward bound traffic lane in dense fog. The fact that the ketch did not participate in the Seattle Traffic Reporting System, that the MOONGLOW was not detected by radar either by the THOMSON or VTS, and that, because her operator was engaged in other tasks, he did not observe the approaching submarine by radar, contributed to this occurrence.Causes And Contributing Factors The collision occurred because the operator of the MOONGLOW did not plot his vessel's position frequently to monitor the vessel's progress along her intended track. As a result, the ketch inadvertently strayed into the outward bound traffic lane in dense fog. The fact that the ketch did not participate in the Seattle Traffic Reporting System, that the MOONGLOW was not detected by radar either by the THOMSON or VTS, and that, because her operator was engaged in other tasks, he did not observe the approaching submarine by radar, contributed to this occurrence.